Defective Activation of ERK in Macrophages Lacking the p50/p105 Subunit of NF-κB Is Responsible for Elevated Expression of IL-12 p40 Observed after Challenge with Helicobacter hepaticus1

Helicobacter hepaticus is an enterohepatic Helicobacter species that induces lower bowel inflammation in susceptible mouse strains, including those lacking the p50/p105 subunit of NF-κB. H. hepaticus-induced colitis is associated with elevated levels of IL-12 p40 expression, and p50/p105-deficient macrophages express higher levels of IL-12 p40 than wild-type macrophages after challenge with H. hepaticus. However, the molecular mechanisms by which the p50/p105 subunit of NF-κB suppresses IL-12 p40 expression have not yet been elucidated. In this study we have demonstrated that H. hepaticus challenge of macrophages induces ERK activation, and this event plays a critical role in inhibiting the ability of H. hepaticus to induce IL-12 p40. Activation of ERK requires both p50/p105 and the MAPK kinase kinase, Tpl-2. Inhibition of the induction of IL-12 p40 by ERK was independent of c-Rel, a known positive regulator of IL-12 p40. Instead, it was linked to the induction of c-Fos, a known inhibitor of IL-12 p40 expression. These results suggest that H. hepaticus induces ERK activation by a pathway dependent upon Tpl-2 and p105, and that activation of ERK inhibits the expression of IL-12 p40 by inducing c-Fos. Thus, a defect in ERK activation could play a pivotal role in the superinduction of IL-12 p40 observed after challenge of macrophages lacking the p50/p105 subunit of NF-κB with H. hepaticus.

I nflammatory microflora that colonize the intestine play a crucial role in the development of inflammatory bowel disease (1)(2)(3). Inappropriate induction of IL-12 expression by these microflora is thought to be one of the central events leading to the initiation of pathological inflammatory responses (4,5). Helicobacter hepaticus (Hh) 3 is an enterohepatic Helicobacter species that can induce lower bowel inflammation in susceptible hosts (6 -9). The development of Hh-induced colitis is closely associated with elevated levels of IL-12 (9,10), and depletion of IL-12 p40 inhibits disease development in a number of Hh-dependent models (8, 10 -12). APCs such as macrophages and dendritic cells are the primary producers of IL-12 p40 after microbial challenge (5). Interestingly, we have found that although Hh is a poor inducer of IL-12 p40 expression in macrophages derived from wild-type (WT) mice that are resistant to colitis, Hh does induce robust expression of IL-12 p40 in macrophages derived from mice that are susceptible to Hh-induced colitis, including mice lacking IL-10 (M. F. Tomczak, S. E. Erdman, A. Davidson, P. R. Nambiar, Y. Y. Wang, D. Luchetti, J. G. Fox, and B. H. Horwitz, manuscript in preparation) or those lacking the p50/p105 subunit of NF-B (10). This suggests that although Hh has the intrinsic ability to induce IL-12 p40 expression, this function is kept in check in WT cells by host regulatory factors. Thus, delineating the interactions between Hh and the host that modulate IL-12 p40 expression may lead to important insights into the regulation of mucosal inflammation.
It has recently been shown that Hh can function as a TLR2 agonist, but not as a TLR4 agonist (13). It has been suggested that the ability of TLR2 agonists to induce IL-12 p40 is inhibited by ERK1/2 (14 -16). In addition, it has been demonstrated that the ability of ERK1/2 to inhibit IL-12 p40 expression depends upon the induction of the immediate-early gene c-fos (14,15), which has been suggested to be a direct inhibitor of IL-12 p40 gene transcription (17). These observations raise the question of whether Hh is able to induce activation of ERK1/2 and its target c-Fos, and furthermore, whether these events have a role in preventing robust expression of IL-12 p40.
We have previously demonstrated a pivotal role for the NF-B subunit p50/p105 in preventing Hh-induced colitis and in limiting the expression of IL-12 p40 from Hh-challenged macrophages (9,10). p50 is one of the five subunits of NF-B (18). It represents the N-terminal fragment of p105 and is thought to be produced by endoproteolytic cleavage. Both p50 and p105 can be detected in most cell types. The C-terminal portion of the p105 molecule has multiple ankyrin repeats, similar to those found in IB molecules, and it has been shown that p105 has IB-like activity. Interestingly, it has previously been shown that the absence of p105 destabilizes the Tpl-2 kinase, which is required for ERK activation by LPS in macrophages (19). As a result, p50/p105-deficient macrophages do not express detectable levels of Tpl-2, and they are defective in ERK activation by LPS (20). p105 binds stoichiometrically to the Tpl-2 kinase, and p105-bound Tpl-2 is stable, but inactive. TLR and death receptor signals promote the release of Tpl-2 from p105, but although active, unbound Tpl-2 is unstable and undergoes rapid degradation via the proteosome. The release of Tpl-2 from p105 is induced by phosphorylation of Tpl-2 at Thr 290 (21,22) and is supported by the signal-dependent degradation of p105 (23,24). These observations have led us to hypothesize that p105 and Tpl-2 may be involved in mediating ERK activation in response to Hh and, if correct, may provide a mechanistic explanation for the higher expression of IL-12 p40 observed after Hh challenge of p50/p105-deficient macrophages.
In this report we demonstrate that Hh induces ERK activation in macrophages, and that activated ERK inhibits IL-12 p40 expression by inducing c-Fos. Furthermore, we are able to show that induction of ERK activity, up-regulation of c-Fos, and inhibition of IL-12 p40 expression after Hh challenge require the presence of both p50/p105 and Tpl-2. Thus, p50/p105-and Tpl-2-dependent ERK activation plays a central role in regulating IL-12 p40 expression after Hh challenge.

Experimental animals
All mice were housed in facilities approved by the Association for the Assessment and Accreditation of Laboratory Animal Care. All experiments were approved by the Harvard Medical Area standing committee on animals. WT, IL-10-deficient mice (2), and p50/p105-deficient mice (25) were backcrossed at least six generations onto the 129S6/SvEvTac background. Tpl-2-deficient mice (19) were backcrossed 10 generations onto the C57BL/6 background. C57BL/6 mice (Taconic Farms) were used as controls for Tpl-2-deficient animals.

Preparation of bone marrow-derived macrophages (BMDM)
Bone marrow was flushed, and single-cell suspensions were cultured overnight in BMDM medium consisting of DMEM, 10% FBS, 5% horse serum, 100 IU/ml penicillin, 100 g/ml streptomycin, 10 mM HEPES, 2 mM L-glutamine, and 10% L cell-conditioned medium, as previously described (10). Nonadherent cells were removed 24 h later by washing with DMEM, and then fresh BMDM medium was added. Four or 5 days later, macrophages were removed from the plates by scraping and replated in 6-well dishes at concentrations of 3-4 ϫ 10 6 /well in medium lacking antibiotics.

Infection with Hh
Hh (strain 3B1; American Type Culture Collection; 51449) was grown on blood agar plates under microaerobic conditions at 37°C. Cultures were examined by Gram stain and phase microscopy for bacteria quality and purity. Bacteria were resuspended in cell culture medium. Concentrations of Hh were assessed by spectrophotometry as previously described (9). Bacteria were added to BMDM at a multiplicity of infection of 250:1, as previously described (10), then centrifuged for 3 min at 600 ϫ g at 12°C to initiate infection. After 1 h, bacteria were killed by adding gentamicin (100 g/ml).

Treatment with UO126
UO126 was obtained from EMD Biosciences and dissolved in DMSO. UO126 or an equivalent amount of vehicle alone was added to a final concentration of 5 M 30 min before challenge of BMDM with Hh.
Immunoblotting BMDM were washed with ice-cold PBS and lysed on the plates in 50 mM Tris-HCl (pH 6.8), 2% SDS, 0.1% bromphenol blue, 10% glycerol, and 100 mM DTT. Lysates were heated for 5 min at 95°C and sheared with a 25-gauge needle. After centrifugation for 10 min at 14,000 rpm at room temperature, supernatants were collected, and protein concentrations were estimated using the Lowry method (Micro kit for total protein; Sigma-Aldrich). Extracts (25 g) were resolved on SDS-polyacrylamide gels and transferred onto Immobilon-P membranes (Millipore) using a TransBlot cell (Bio-Rad). Membranes were probed with polyclonal Abs against ERK, phospho-ERK, MEK, phospho-MEK (Cell Signaling Technology), and c-Fos (Upstate Biotechnology). After incubation with HRP-conjugated antirabbit Abs, reactive proteins were visualized with Supersignal West Pico chemiluminescent substrate (Pierce). The densities of the bands were quantified by histogram analysis using Adobe Photoshop, and the ratios of phosphorylated protein to total protein were calculated.

RNase protection
To collect RNA from BMDM, cells were lysed on the plates with 1 ml of TriReagent (Molecular Research Center). RNA was isolated according to the manufacturer's instructions. RNase protection analyses were performed on 5-10 g of total RNA using RiboQuant MultiProbe Template sets (BD Pharmingen). The intensities of the protected fragments were quantified by phosphorimager analysis, and the relative expression of IL-12 p40 to GAPDH was calculated. To allow normalization between experiments, the maximal relative expression of IL-12 p40 in each experiment was calculated, and expression in other samples is represented as a percentage of this maximal value.

Statistical analysis
Unpaired two-tailed t test was used to compare the gene expression and ELISA data. The differences were considered statistically significant at p Ͻ 0.05.

Hh induces ERK activation that is dependent upon p50/p105
To determine whether Hh induces ERK1/2 activation, BMDM were challenged with live Hh and induction of ERK phosphorylation and phosphorylation of the ERK-specific kinase MEK1/2 were monitored by Western blotting with phospho-specific Abs (Fig. 1a). Challenge with Hh induced rapid phosphorylation of ERK1/2 and MEK1/2, demonstrating that Hh is a potent inducer of the ERK pathway. Interestingly, there was a severe defect in the phosphorylation of both ERK1/2 and MEK1/2 in BMDM harvested from p50/p105-deficient mice, suggesting that p50/p105 is necessary for Hh-induced ERK activation (Fig. 1a).
Previous results suggested that the presence of p105 is necessary to stabilize Tpl-2 kinase (20 -24), a MEK kinase involved in ERK activation after LPS stimulation (19). To determine whether Tpl-2 was necessary for Hh-induced ERK activation, BMDM were harvested from Tpl-2-deficient mice, and MEK and ERK activation was evaluated after challenge with Hh (Fig. 1b). Tpl-2-deficient BMDM exhibited a marked defect in Hh-induced MEK and ERK phosphorylation, demonstrating that Tpl-2 is required for Hh-induced MEK and ERK activation.
IL-10 is a potent inhibitor of IL-12 p40 in many situations (26,27), and we have recently demonstrated that IL-10 inhibits IL-12 p40 expression after Hh challenge (M. F. Tomczak, S. E. Erdman, A. Davidson, P. R. Nambiar, Y. Y. Wang, D. Luchetti, J. G. Fox, and B. H. Horwitz, manuscript in preparation). Although there is abundant evidence that STAT-3 activation mediates the inhibitory functions of IL-10 (26, 28), it has not been determined whether the absence of IL-10 affects the activation of ERK or potential inhibitory pathways downstream of ERK. Interestingly, it has been suggested that IL-10 induces the expression of Tpl-2 mRNA by a STAT-3-dependent mechanism (26), raising the possibility that defective Tpl-2 expression in the absence of IL-10 could lead to a defect in ERK activation. To directly determine whether IL-10 is necessary for efficient ERK activation, BMDM harvested from IL-10-deficient mice were challenged with Hh, and phosphorylation of ERK1/2 and MEK1/2 was evaluated (see Fig. 1a). There were no defects observed in the induction of ERK1/2 or MEK1/2 phosphorylation in WT and IL-10-deficient BMDM. Thus, the ability of Hh to induce ERK activation is independent of IL-10.

Hh-induced ERK activation inhibits the expression of IL-12 p40
We have previously demonstrated that Hh induces elevated levels of IL-12 p40 in p50/p105-deficient BMDM compared with levels induced in WT BMDM (10). Consistent with previous results, RNase protection analysis confirmed that Hh induced higher levels of IL-12 p40 mRNA at 4, 6, 8, and 12 h in p50/p105-deficient BMDM than in WT BMDM (Fig. 2a), and this was accompanied by higher levels of IL-12 p40 secretion at both 8 and 12 h in p50/p105-deficient BMDM than in WT BMDM (Fig. 2b). To determine whether the absence of ERK activity observed in p50/ p105-deficient cells was responsible for the observed increases in Hh-induced IL-12 p40 expression, WT BMDM were treated with UO126, a selective inhibitor of ERK1/2 activation (29), for 30 min before stimulation with Hh. Treatment with UO126 effectively inhibited the ability of Hh to induce ERK phosphorylation (Fig. 3a). In addition, UO126-treated BMDM expressed higher levels of IL-12 p40 mRNA (Fig. 3b) and protein (Fig. 3c) than mock-treated BMDM after Hh challenge, and in both cases, this difference reached statistical significance at 8 and 12 h. Furthermore, Tpl-2deficient BMDM also expressed higher levels of IL-12 p40 than control BMDM after Hh treatment (Fig. 3, e and f). Elevated levels of IL-12 p40 led to the production of IL-12 p70 (Fig. 3, d and g). Taken together, these results strongly suggest that Hh-induced ERK activity inhibits the expression of IL-12 p40, and that this, in turn, limits secretion of the bioactive species IL-12 p70.

ERK can inhibit Hh-induced expression of IL-12 p40 independently of changes in IL-10 secretion
Several previous studies have shown that ERK activity enhances the secretion of IL-10 (16,30). These studies suggest that ERK may inhibit the expression of IL-12 p40 by inducing IL-10. However, it has also been suggested that ERK may inhibit the expression of IL-12 p40 by mechanisms independent of IL-10 (16). BMDM treated with UO126 secreted lower levels of IL-10 after challenge with Hh than mock-treated BMDM (Fig. 4a). Furthermore, both p50/p105-deficient and Tpl-2-deficient BMDM secreted lower levels of IL-10 than matched control cells (data not shown). Thus, consistent with the results of previous experiments using other TLR2 ligands, induction of ERK activity after Hh challenge enhances the expression of IL-10. Although, lower levels of IL-10 expression are likely to be an important factor leading to elevated expression of IL-12 p40 in the absence of ERK activity, it has also been proposed that ERK may play a more direct role in regulating the expression of IL-12 p40. Therefore, to definitively determine whether ERK activity has the ability to inhibit IL-12 p40 expression independently of the induction of IL-10, we compared Hh-induced IL-12 p40 expression in IL-10-deficient BMDM treated, or not, with UO126 in the presence of a fixed concentration of exogenous IL-10 (Fig. 4b). Interestingly, in the presence of IL-10 (1.0 ng/ml), Hh induced higher expression of IL-12 p40 in UO126-treated, IL-10-deficient BMDM than in mock-treated, IL-10-deficient BMDM at both 8 and 12 h. These results imply that Hh-induced ERK activity inhibits IL-12 p40 expression in a fashion that is at least in part independent of IL-10.

Hh-induced c-Fos expression depends on the activation of ERK
It had been suggested previously that the induction of c-Fos after TLR2 ligation depends on the activation of ERK, and that c-Fos can inhibit the expression of IL-12 p40 (14, 15, 17). To determine FIGURE 1. ERK activation after Hh challenge requires p50/p105 and Tpl-2. a, Immunoblots of total cell extracts prepared from pools of BMDM derived from four to eight WT, p50/p105-deficient (p50 Ϫ/Ϫ ), or IL-10deficient (IL-10 Ϫ/Ϫ ) animals as indicated. Before harvest, BMDM were stimulated with Hh for the time in minutes shown below, and blots were probed with Abs shown on the right. Bands were quantified by digital imaging, and the relative ratio (ratio) of phosphorylated to total protein is shown above each phosphorylated band. Results are indicative of at least three independent experiments. b, Immunoblots of total cell extracts from WT or Tpl-2-deficient (Tpl2 Ϫ/Ϫ ) BMDM. Before harvest, BMDM were stimulated with Hh for the time in minutes shown below, and blots were probed with the Abs shown on the right. The relative ratio (ratio) of phosphorylated to total protein is shown above each phosphorylated band. Results are representative of two independent experiments. whether Hh regulates c-Fos expression, c-Fos levels were measured in total cell extracts by Western blotting (Fig. 5). Hh induced the expression of c-Fos beginning at 60 min and extending up to 180 min after Hh challenge. Induction of c-Fos was inhibited by the presence of UO126, as expected (Fig. 5, top panel). Furthermore, there was a defect in the induction of c-Fos after Hh challenge in both p50/p105-deficient BMDM (Fig. 5, middle panel) and Tpl-2-deficient BMDM (Fig. 5, bottom panel). There was no difference in the induction of c-Fos in the absence of IL-10 (Fig. 5,   middle panel). Thus, the induction of c-Fos closely correlates with the activation of ERK, but not with the expression of IL-10.

ERK-mediated inhibition of IL-12 p40 expression does not depend on the inhibition of c-Rel
It had been demonstrated previously that the induction of IL-12 p40 expression after challenge of macrophages with the combination of LPS and IFN-␥ depends upon the presence of c-Rel (31). Furthermore, it had been suggested that the ability of IL-10 to FIGURE 3. Activation of ERK suppresses the expression of IL-12 p40 after Hh challenge. a, Immunoblots of total cell extracts prepared from pools of BMDM derived from four to eight WT mice. Before harvest, BMDM were stimulated with Hh in the presence (UO126) or the absence (Mock) of UO126 for the time in minutes shown below, and blots were probed with the Abs shown on the right. The relative ratio of phosphorylated to total protein is shown above each phosphorylated band. The results are indicative of at least three independent experiments. b, Percent maximal expression of IL-12 p40 mRNA relative to GAPDH in WT BMDM challenged with Hh in the presence (Ⅺ) or the absence (f) of UO126, as determined by phosphor imager analysis of RPA. Each data point represents the mean value derived from three independent experiments analyzing BMDM pooled from four to eight individual animals per experiment. The SEM is shown. c, Secretion of IL-12 p40 into the culture medium of WT BMDM challenged with Hh in the presence (Ⅺ) or the absence (f) of UO126 at 8 and 12 h after stimulation, as determined by ELISA. Each data point represents the mean value of IL-12 p40 observed in the culture medium in two independent experiments. The SEM is shown. d, Secretion of IL-12 p70 into the culture medium of WT BMDM challenged with Hh in the presence (Ⅺ) or the absence (f) of UO126 at 8 and 12 h after stimulation, as determined by ELISA. Each data point represents the mean value of IL-12 p70 observed in the culture medium in two independent experiments. The SEM is shown. e, Percent maximal expression of IL-12 p40 mRNA relative to GAPDH in WT (f) and Tpl-2-deficient (Ⅺ) BMDM challenged with Hh, as determined by phosphorimager analysis of RPA. Each data point represents the mean value derived from two independent experiments analyzing BMDM pooled from four to eight individual animals per experiment. The SEM is shown. f, Secretion of IL-12 p40 into the culture medium of WT (f) and Tpl-2-deficient (Ⅺ) BMDM challenged with Hh at 8 and 12 h after stimulation, as determined by ELISA. Each data point represents the mean value of IL-12 p40 observed in the medium in two independent experiments. The SEM is shown. g, Secretion of IL-12 p70 into the culture medium of WT (f) and Tpl-2-deficient (Ⅺ) BMDM challenged with Hh at 8 and 12 h after stimulation, as determined by ELISA. Each data point represents the mean value of IL-12 p70 observed in the medium in two independent experiments. The SEM is shown. ‫,ء‬ p Ͻ 0.05 (between groups). inhibit IL-12 p40 expression depends on inhibition of the nuclear translocation of c-Rel (32). We, therefore, considered the possibility that the induction of IL-12 p40 by Hh may also depend on c-Rel. We also hypothesized that ERK might interfere with IL-12 p40 expression by inhibiting c-Rel function. To test this hypothesis, we compared the abilities of Hh to induce IL-12 p40 expression in WT and c-Rel-deficient macrophages (Fig. 6a). Hh induced lower levels of IL-12 p40 in c-Rel-deficient BMDM than in WT BMDM. This demonstrates that the ability of Hh to induce IL-12 p40 depends at least in part on the presence of c-Rel. Next, we evaluated the effect of UO126 treatment on the ability of c-Reldeficient BMDM to express IL-12 p40. We observed considerably higher levels of IL-12 p40 expression in c-Rel-deficient BMDM treated with UO126 than in those treated with vehicle control (Fig.  6b). These results indicate that inhibition of ERK activity leads to increased expression of IL-12 p40, even in the absence of c-Rel, suggesting that the ability of ERK to inhibit IL-12 p40 expression does not depend on inhibition of c-Rel function.

Discussion
In this study we have demonstrated that Hh challenge of BMDM induces ERK activation, and that this event requires both p50/p105 and the MEK kinase, Tpl-2. ERK activity appears to play a critical role in inhibiting the ability of Hh to induce IL-12 p40, because p50/p105-deficient BMDM, Tpl-2-deficient BMDM, and UO126treated WT BMDM all express higher levels of IL-12 p40 after Hh challenge than the appropriate controls. Inhibition of IL-12 p40 by ERK is at least in part independent of modulation of IL-10 expression, because IL-10-deficient cells treated with the ERK inhibitor UO126 produced higher levels of IL-12 p40 after Hh stimulation than untreated IL-10-deficient cells. Furthermore, inhibition by ERK was also independent of modulation of c-Rel function, because the ability of UO126 to enhance Hh-induced IL-12 p40 expression was not abrogated in c-Rel-deficient BMDM. However, there was a close correlation between inhibition of Hh-induced IL-12 p40 expression and induction of c-Fos, a known inhibitor of IL-12 p40 expression, suggesting that c-Fos inhibits the ability of Hh to induce IL-12 p40 expression. These results suggest the model shown in Fig. 7. In this model, Hh challenge induces ERK activation by a pathway dependent upon Tpl-2 and p105. Once activated, ERK induces increased levels of c-Fos protein that, in turn, inhibit the expression of IL-12 p40.
It has been demonstrated that Hh is a TLR2 ligand (13). However, the pathways that lead to ERK activation after Hh challenge remain incompletely defined. After ligation of receptor tyrosine kinases such as c-Kit, sequential activation of the MEK kinase Raf and MEK1 and -2 lead to activation of ERK1/2 (33). In some cells it has been demonstrated that this process also requires activation of the PI3K/AKT pathway (34,35). It has been reported that activation of ERK1/2 after TLR2 ligation is also mediated by Raf in RAW 264.7 macrophages (36). However, other studies have conclusively shown that ERK activation in primary macrophages by LPS depends upon the activation of the MEK kinase, Tpl-2 (19). Tpl-2 was originally identified as a proto-oncogene activated by provirus integration in Moloney murine leukemia virus-induced rodent T cell lymphomas and mouse mammary tumor virus-induced mammary carcinomas (37,38). The observation that Tpl-2 is required for Hh-induced ERK activity suggests that Tpl-2 may be necessary for ERK activation after TLR2 ligation as well as TLR4 ligation, probably due to its ability to induce MEK phosphorylation.
Previous results have demonstrated that the decreased Tpl-2 signaling observed after LPS stimulation of p50/p105-deficient macrophages is not due to decreased expression of the Tpl-2 gene,  because similar levels of Tpl-2 mRNA were found in WT and p50/p105-deficient cells. Instead, p105 binds to and stabilizes Tpl-2 (20 -24). Our observation that there is a defect in the ability of Hh to induce ERK activation in p50/p105-deficient BMDM as well as Tpl-2-deficient BMDM suggests that the ability of p105 to stabilize Tpl-2 may be essential for ERK activation in response to Hh as well as LPS. Furthermore, these results raise the question of whether the primary inhibitory function of p50/p105 on IL-12 p40 expression is the result of its role in activating ERK. Previous studies have suggested that interaction of p50 homodimers with DNA may have a direct inhibitory role on gene transcription (39), possibly mediated by the ability of p50 homodimers to recruit histone deacetylases to promoter complexes (40), a process associated with histone deacetylation and inhibition of gene transcription. It has also been suggested that the ability of p50 homodimers to recruit Bcl-3 may play an important role in inhibiting proinflammatory genes, including TNF (41). Whether these other potential functions of p50/p105 could also be involved in inhibitory processes that work in parallel with ERK will require additional investigation. It has been demonstrated that the ability of ERK to suppress IL-12 p40 depends upon increased secretion of IL-10 (16,30), and we have verified these previous findings. However, it has also been suggested that ERK-mediated suppression of IL-12 p40 can pro-ceed in a manner that is at least partially independent of IL-10 modulation (16). In this report we have demonstrated that inhibition of ERK increases the expression of IL-12 p40 in IL-10-deficient BMDM treated with a fixed amount of exogenous IL-10. Furthermore, we have observed similar levels of STAT-3 phosphorylation after Hh stimulation of U0126-treated and untreated BMDM (data not shown), indicating that ERK inhibition does not interfere with the ability of IL-10 to induce STAT-3. Taken together, these results suggest that ERK inhibits the expression of IL-12 p40 in a fashion that is at least in part independent of the induction of IL-10. Thus, both IL-10-dependent and -independent pathways may be involved in ERK-mediated suppression of IL-12 p40.
Members of the Jun family heterodimerize with c-Fos and related family members to form transcription factors of the AP-1 family (42). AP-1 binding sites have been found in many genes involved in immune and inflammatory responses, including IL-12 p40 (43). Although in most situations Fos family members have been associated with the induction of transcription (42), it has been previously suggested that c-Fos may have a role in inhibiting the expression of IL-12 p40 (14,15,17). It has been demonstrated that ERK and an ERK-activated kinase, ribosomal S6 kinase, can induce sequential phopshorylations at multiple sites within c-Fos after stimulation of cells with epidermal growth factor, and that this event significantly enhances protein stability, contributing to the increased abundance of c-Fos observed at later time points after stimulation (44,45). In this study we have found that activation of ERK potently influences the ability of Hh to induce c-Fos. Because we have not observed changes in c-Fos mRNA levels after Hh stimulation (data not shown), these observations suggest that induction of c-Fos after Hh stimulation may be the result of enhanced protein stability. Furthermore, our observations that defective ERK activation and elevated expression of IL-12 p40 are tightly correlated with a failure to induce c-Fos suggest that the induction of c-Fos is critical to inhibiting IL-12 p40 expression after Hh stimulation. Thus, these studies confirm a key role for c-Fos in inhibiting IL-12 p40 expression after TLR-2 ligation (14,15).
Previous studies have implied that NF-B complexes containing c-Rel are required for the induction of IL-12 p40 gene transcription after stimulation of macrophages with LPS in combination with IFN-␥ (31). Consistent with these results, we have demonstrated that Hh induces lower levels of IL-12 p40 in c-Rel-deficient macrophages than in WT macrophages. However, inhibition of ERK activity in c-Rel-deficient macrophages still augmented the ability  of Hh to induce IL-12 p40 expression, indicating that ERK-mediated inhibition of IL-12 p40 expression is unlikely to depend on interfering with the function of c-Rel. These results suggest that c-Rel and p50/p105 appear to have opposite effects on regulating the ability of Hh to induce IL-12 p40 expression. Although the inhibitory effects of p50/p105 appear to be mediated at least in part through its ability to induce ERK activation, whether the activity of c-Rel is the result of direct interaction with the NF-B half-site present in the IL-12 p40 promoter or, alternatively, is the result of an indirect mechanism requires additional study.
It has been suggested that the regulation of Hh-induced IL-12 p40 expression is critical to preventing Hh-induced colitis (8,10,12), and modulation of IL-12 p40 expression may also have an important role in preventing inflammatory bowel disease in humans (46). The observation that ERK has a central role in regulating microflora-induced IL-12 p40 expression in BMDM raises the question of whether ERK also has an important role in inhibiting microflora-induced colitis in animal models. Interestingly, although mice lacking p50/p105 are susceptible to colitis induced by Hh, mice that are heterozygously deficient at the NF-B p65 locus in addition to lacking p50/p105 (3X mice) develop more severe colitis after Hh challenge than mice lacking p50/p105 alone (9). Because the defect in Hh-induced ERK activation is quite severe in BMDM derived from mice lacking p50/p105 alone and is no more severe in BMDM derived from 3X mice (data not shown), it seems unlikely that an additional defect in ERK activation can explain the increased sensitivity of 3X mice to Hhinduced colitis. Nonetheless, because mice that are heterozygous for p65 and WT for p50/p105 are not sensitized to the development of colitis (9), it is clear that p50/p105 plays a central role in protecting mice from this disease. Thus, because we have shown that defective ERK activation in p50/p105-deficient BMDM is an important factor leading to dysregulated expression of IL-12 p40 after Hh challenge, it seems potentially possible that this defect could, in fact, play an important role in inhibiting microflora-induced colitis in animal models. Therefore, we suggest that directly assessing the role of ERK in inhibiting microflora-induced colitis in mice and, also, determining whether alterations in this pathway could have a role in the development of human inflammatory bowel diseases will be important topics for future investigation.

Disclosures
The authors have no financial conflict of interest.